For a couple of months last summer, neuroscientist Justin Wood filmed himself aiming rocks at monkeys. Technically, he was only pretending to hit the monkeys. He just wanted to see how they'd react, in a bid to pin down the neurologic underpinnings of a uniquely human trait: throwing.
Early every morning, under a sky streaked with the colors of dawn, the Harvard graduate student in cognitive neuroscience climbed aboard a small boat from the Puerto Rican city of Punta Santiago, and took a 10-minute trip to nearby Cayo Santiago. The island is also known as Monkey Island, in honor of the 1,000 or so macaques that reside among its scattered palm trees and gentle terrain. They are the offspring of monkeys first imported for research purposes to the island from India in the 1940s.
Wood's project was a somewhat lonely one - and thankless, given that the monkeys likely didn't appreciate him taking aim at their foreheads. "Once I was on the island, I would walk around by myself with a video camera, looking for a group of monkeys," Wood explains. "Then I would set up the video camera and show them that I was holding a rock and perform different throwing actions and see whether they flinched."
Macaques are physically incapable of targeting an object in a specific direction, due to their shoulder, hip, and buttock physiology. (Although the image of monkeys tossing "feces at you at zoos" may be familiar to many, the monkeys are flinging it in a general direction, rather than aiming at something specific, Wood explains.)
All told, Wood subjected the monkeys to 12 different actions designed to study how well they grasped the principles of throwing. For instance, he'd pull his arm back and rotate it forward until it was in line with his body, without extending it forward; other times he'd just extend his arm without bringing it back first.
When he and other researchers reviewed the footage, it revealed that 85% of the monkeys got out of the way when Wood mimed a complete under-arm or over-arm throw. On the other hand, few were worried by any partial throws, or throws that seemed to lack the requisite speed or torque to be a threat.
So if the monkeys know that a throw is aimed at them and could hurt them, it suggests they can reason about throwing, even if they can't do it themselves. This finding may mean that the physical and psychological aspects of throwing evolved separately, Wood says. (see "Do Chimps Have Culture?")
His results also directly contradict a theory that philosophers and neuroscientists have proposed: We must be able to perform an action in order to understand it. This theory is supported by the existence of mirror neurons, which fire when an animal both performs and observes an action. "Our study challenges this position, by showing that despite the fact that rhesus monkeys cannot throw themselves, they are perfectly capable of reasoning about throwing actions performed by others," Wood says.
Melvyn Goodale from the University of Western Ontario agrees that the study shows that mirror neurons can't explain everything. "The idea of the mirror neuron has become a kind of touchstone for everything from social cognition to intention understanding," he says. "The concept has become overextended."
Wood's principal investigator, Marc Hauser, who has been studying the Santiago monkeys for more than 20 years, says that by separating the idea of throwing from the act itself, Wood's research fits into a much larger picture, one that includes tool use, mathematics, language, and other areas. "Animals have far greater competence to understand the world than they do to create," Hauser says. "Thus, animals can understand what a good tool is, but have little capacity to create complex tools. Animals have the capacity to extract grammatical structures from sounds in the environment, even though they don't produce grammatically structured sounds. And now, with this study, animals understand the function and kinematics of throwing without being able to throw."